From: "Vincent Stehlé" <vincent.stehle@arm.com>
To: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Cc: Heinrich Schuchardt <xypron.glpk@gmx.de>,
Ilias Apalodimas <ilias.apalodimas@linaro.org>,
Tom Rini <trini@konsulko.com>,
Jose Marinho <jose.marinho@arm.com>,
u-boot@lists.denx.de
Subject: Re: [PATCH] efi_loader: fix ecpt size computation
Date: Wed, 11 Feb 2026 12:47:53 +0100 [thread overview]
Message-ID: <aYxsaXdISybHOQTq@debian> (raw)
In-Reply-To: <a386de9e-1dc9-4b50-a86c-bb8dfd845eef@gmx.de>
On Wed, Feb 11, 2026 at 10:56:30AM +0100, Heinrich Schuchardt wrote:
> On 2/11/26 10:33, Vincent Stehlé wrote:
> > The size of the memory allocated for the EFI Conformance Profiles Table is
> > computed with `num_entries' always equal to zero, which is incorrect when
> > CONFIG_EFI_EBBR_2_1_CONFORMANCE is enabled.
> >
> > This can be verified by allocating the ECPT memory with malloc() instead of
> > efi_allocate_pool(), building u-boot with sandbox_defconfig and
> > CONFIG_VALGRIND=y, and by finally running the following command:
> >
> > valgrind --suppressions=scripts/u-boot.supp \
> > ./u-boot -T -c 'efidebug tables'
> >
> > Fix this by using an array of the supported profiles GUIDs instead, which
> > should also be easier to extend in the future.
>
> Thank you Vincent for this patch.
Hi Heinrich,
Thanks for reviewing.
> Maybe mention in the commit message that U-Boot should publish all supported
> EBBR revisions.
Ok, will do.
>
> >
> > Fixes: 6b92c1735205 ("efi: Create ECPT table")
> > Signed-off-by: Vincent Stehlé <vincent.stehle@arm.com>
> > Cc: Heinrich Schuchardt <xypron.glpk@gmx.de>
> > Cc: Ilias Apalodimas <ilias.apalodimas@linaro.org>
> > Cc: Tom Rini <trini@konsulko.com>
> > Cc: Jose Marinho <jose.marinho@arm.com>
> > ---
> > lib/efi_loader/efi_conformance.c | 18 +++++++++---------
> > 1 file changed, 9 insertions(+), 9 deletions(-)
> >
> > diff --git a/lib/efi_loader/efi_conformance.c b/lib/efi_loader/efi_conformance.c
> > index 2bae93a94bd..9768b5ae824 100644
> > --- a/lib/efi_loader/efi_conformance.c
> > +++ b/lib/efi_loader/efi_conformance.c
> > @@ -13,8 +13,12 @@
> > #include <malloc.h>
> > static const efi_guid_t efi_ecpt_guid = EFI_CONFORMANCE_PROFILES_TABLE_GUID;
> > -static const efi_guid_t efi_ebbr_2_1_guid =
> > - EFI_CONFORMANCE_PROFILE_EBBR_2_1_GUID;
> > +
> > +static const efi_guid_t profiles[] = {
>
> As profiles[] is not used outside efi_ecpt_register(), there is no need to
> make it a static variable.
I am not sure why a "global" symbol would be better than a static one, but if
you like this solution we could make the profiles[] array extern.
Then I guess we should rename it with a much longer name to avoid namespace
"pollution":
const efi_guid_t efi_conformance_profiles[] = {
...
All the sizeof() and ARRAY_SIZE() would have to follow, too.
Alternatively, we could move the array declaration inside the
efi_ecpt_register() function completely:
efi_status_t efi_ecpt_register(void)
{
struct efi_conformance_profiles_table *ecpt;
efi_status_t ret;
size_t ecpt_size;
static const efi_guid_t profiles[] = {
#if CONFIG_IS_ENABLED(EFI_EBBR_2_1_CONFORMANCE)
EFI_CONFORMANCE_PROFILE_EBBR_2_1_GUID,
#endif
};
...
This is still static here, but for storage duration not linkage. One advantage
is that the name would not have to change. I am not a big fan of the #if in
function scope, but if you prefer this solution this is fine as well.
Best regards,
Vincent.
>
> Otherwise looks good to me.
>
> Best regards
>
> Heinrich
>
>
> > +#if CONFIG_IS_ENABLED(EFI_EBBR_2_1_CONFORMANCE)
> > + EFI_CONFORMANCE_PROFILE_EBBR_2_1_GUID,
> > +#endif
> > +};
> > /**
> > * efi_ecpt_register() - Install the ECPT system table.
> > @@ -23,12 +27,11 @@ static const efi_guid_t efi_ebbr_2_1_guid =
> > */
> > efi_status_t efi_ecpt_register(void)
> > {
> > - u16 num_entries = 0;
> > struct efi_conformance_profiles_table *ecpt;
> > efi_status_t ret;
> > size_t ecpt_size;
> > - ecpt_size = num_entries * sizeof(efi_guid_t)
> > + ecpt_size = sizeof(profiles)
> > + sizeof(struct efi_conformance_profiles_table);
> > ret = efi_allocate_pool(EFI_BOOT_SERVICES_DATA, ecpt_size,
> > (void **)&ecpt);
> > @@ -39,12 +42,9 @@ efi_status_t efi_ecpt_register(void)
> > return ret;
> > }
> > - if (CONFIG_IS_ENABLED(EFI_EBBR_2_1_CONFORMANCE))
> > - guidcpy(&ecpt->conformance_profiles[num_entries++],
> > - &efi_ebbr_2_1_guid);
> > -
> > + memcpy(ecpt->conformance_profiles, profiles, sizeof(profiles));
> > ecpt->version = EFI_CONFORMANCE_PROFILES_TABLE_VERSION;
> > - ecpt->number_of_profiles = num_entries;
> > + ecpt->number_of_profiles = ARRAY_SIZE(profiles);
> > /* Install the ECPT in the system configuration table. */
> > ret = efi_install_configuration_table(&efi_ecpt_guid, (void *)ecpt);
>
next prev parent reply other threads:[~2026-02-11 11:48 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-02-11 9:33 [PATCH] efi_loader: fix ecpt size computation Vincent Stehlé
2026-02-11 9:56 ` Heinrich Schuchardt
2026-02-11 11:47 ` Vincent Stehlé [this message]
2026-02-11 11:50 ` Heinrich Schuchardt
2026-02-11 12:25 ` Vincent Stehlé
2026-02-12 14:40 ` [PATCH v2] " Vincent Stehlé
2026-02-15 7:30 ` Heinrich Schuchardt
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=aYxsaXdISybHOQTq@debian \
--to=vincent.stehle@arm.com \
--cc=heinrich.schuchardt@canonical.com \
--cc=ilias.apalodimas@linaro.org \
--cc=jose.marinho@arm.com \
--cc=trini@konsulko.com \
--cc=u-boot@lists.denx.de \
--cc=xypron.glpk@gmx.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox